I am a bit insecure concerning your phrasing. Are you asking for detecion limits of 10-20 ppb? accuracy is asking for precision and in environmental samples you would usually assume you get something between 5-20 standard deviation.

with electron impact you have a chance to see about 1 pg injected onto the column. with a good trap you can get maybe 10 mL air onto the column directly so in ppbs (V/V) that would be 100 ppb. I guess you have to trap the N2O offline with a chemical trap upfront

You can also use an ftir with a long path cell. A cell with an optical path of 10 meters should be enough to be anle to measure 10-20 ppbs and the procedure is very simple...